- This report dissects Jim Cramer’s career as a fund manager and media personality in order to understand his influence on the market.
- In his hedge fund manager days, we find that many of the stocks he picked outperformed the market in 1999, albeit over a longer timeline.
- Cramer’s Mad Money stock picks were heavily skewed on the buy side. Digging deeper into his recommendations reveals sector-specific preferences.
- A “buy and hold” strategy on Cramer’s Mad Money buy recommendations since 2016 is unlikely to beat the market in the long run.
- Cramer’s show had a detectable impact on stock daily returns. More interestingly, we found the once lauded phenomenon known as the “Cramer Bounce” seems to now be a “Cramer Dive”.
- This decrease in immediate returns following Cramer’s recommendations may be due to a change in attitude by investment professionals.
- This report is not intended for financial investment advice in any form.
- The interpretation and calculation of stock returns. Daily returns are calculated on the same day open and close prices of the underlying stock. This calculation is more applicable to retail traders who usually do not trade outside official trading hours (930 am -4pm US ET), hence cannot front-run others upon hearing the Cramer news (6 pm US ET). The return will vary significantly if calculated otherwise, thus skewing the interpretation for retail traders.
About This Report
This report will cover Jim Cramer’s career from Harvard to Mad Money to his most recently announced partnership. By understanding where he’s been, what he’s accomplished, his behaviour and his influence using data, we hope to help retail traders understand Cramer a little better.
Part I: The Early Years of James J. Cramer
For most folks, graduating from Harvard Law School is a career defining moment. Not for Jim Cramer. Around his graduation in 1984, his standing among Harvard alumni (owing to his charismatic personality) and stock picking reputation had connected him with something better, a $500,000 token from publisher Martin Petrez. Cramer made $150,000 in profit in the following two years or a 30% return that earned Petrez’s favor and ignited Cramer’s career. Interestingly, the S&P 500 had similar returns during those years, yielding 36.39% (31.7% if adjusting for inflation). Regardless, the pair eventually founded TheStreet.com in 1996.
But back to the 80s! Jim Cramer had a short stint at Goldman Sachs before launching his own private hedge fund Cramer & Co. with Jeff Berkowtiz in 1987 (early investors included Petrez). The $265 million fund was dwarfed in size by more famous ones, like George Soros's $7.3 billion Quantum Fund, but the fund outshone its competitors in popularity due to Cramer’s media savviness. Unfortunately for us, there are no publicly accessible records of Cramer & Co.’s portfolio holdings or fund performance. Cramer’s own statements allude to how his fund was largely immune to the October 19th “Black Monday” in 1987 . The only year with negative returns was 1998 when the S&P 500 index returned 29%. This news was apparently leaked to the public by Mr. Petrez, leading to investors withdrawing in 1999 and a public feud on the eve of TheStreet IPO. Cramer left his role at Cramer & Co. in 2001. By this point, the fund claims it had averaged a 24% annual return.
The Hedge Fund Manager
There has never been an accurate public record of Cramer’s portfolio performance during his 14 years as a hedge fund manager. However, there have been numerous publications investigating Cramer’s stock recommendations made as a media personality and these have been used as a surrogate metric for his financial wit.
To approximate Cramer & Co.’s holdings and performance, we tracked public data from the SEC’s 13F filings. Here we found eight quarterly filings from Cramer & Co. for the periods between 1999 and early 2000. 13F filings are mandatory quarterly reports filed by institutional investment managers with at least $100 million in assets under management. They disclose equity holdings and can provide insights into what the smart money is doing in the market. Unfortunately, these filings don’t include sales data so generating an accurate portfolio isn’t possible. Many of the stocks from these Cramer & Co. filings are now defunct. We tried our best to sub-sample some NYSE and NASDAQ traded stocks with price data to get a peek into his stock holding performances for 1999 as shown in Figure 1. Although this was only a portion of his portfolio for 1999 we have to agree that Cramer had some really good picks for his own fund. Thanks to the influx of IPOs that year, Cramer’s holdings had largely beat the market bench S&P500 (^GSPC).
Figure 1. Cramer’s 1999-2000 stock performances
We’ve briefly touched on Cramer’s early career highlights to understand his hedge fund management experience. Now, let’s dive into Cramer’s ability to pick stocks and his influence on the market. We’ll use Mad Money’s stock recommendations, market data and social media data to measure Cramer’s influencer impact.
Part II: Mad for Money
What stocks did Cramer pick?
Cramer has made 24,549 stock recommendations on Mad Money between April 2016 to 2021, this averages out to about 10+ stock mentions a day. Unfortunately, details on stock picks prior to 2016 were not easily accessible, so we’ll keep the analyses to this time period. But note that Mad Money first aired in 2005 and so you can imagine how many stock recommendations he’s made in his career.
This analysis used the following data inclusion criteria:
- Mentioned stocks must have clear Buy or Sell signal from Cramer
- Stocks ignored if in Lighting Round segment
- Stocks must be traded on NYSE
Following these filters, a total of 8,530 stock recommendations (1,058 individual tickers) were included in this part of the report. Figure 2 uses a treemap to visualize Cramer’s buy and sell recommendations per sector. Buy side recommendations are colored in Green, while Sell side recommendations are in Red. The “sell” here does not mean taking a short position. Each sector square is sized relative to the proportion of sector recommendations in Cramer’s list. The majority of Cramer’s stock picks were buy recommendations. In fact, 8070 stock mentions (94.6%) were recommending viewers to buy.
Figure 2. Overview of Cramer’s Buy and Sell recommendations
Findings in Figure 1 were consistent with previous publications that looked at Cramer’s early 2000s stock picks, which also revealed a heavy skew on the buy side.
To see if there were any sector preferences in Cramer’s stock picks on the buy and sell side, stock mentions were normalized to the total counts on each size in Figure 3. The comparison reiterates Cramer’s bias for tech and consumer cyclical sectors first observed in Figure 2. We also see a preference for consumer cyclical stocks on the sell side.
Figure 3. Sector distribution of Cramer’s stock picks between 2016-2021
Comparing Cramer’s sector recommendations from 2016 to 2020, we see that different sectors are recommended depending on the year. Racing bar graphs in Figures 4a and 4b animate these changes in sector stock picks per year on both the buy and sell side.
Figure 4a. Stock Sector Buy Recommendations by Year
On the buy side, we can see Cramer’s preference for tech and consumer cyclical remains largely consistent. The rest of the sectors mentioned on the buy side remained relatively stable over the years, except for the drop observed in consumer defense from 2019 to 2020, which could’ve been a reaction to the pandemic.
Figure 4b. Stocks Sector Sell Recommendations by Year
On the sell side, consumer cyclical remained on top year over year, while all other sectors shuffled about.
Different attitudes towards different stocks
So far, we’ve observed Cramer’s preference for specific sectors. How about a preference for specific stock tickers? In Figures 5a and 5b, the relative frequency of each stock ticker was mapped to the sector and buy/sell recommendation. Here we show the Top 10 tickers recommended per sector.
Figure 5a. Top 10 Stock Picks per Sector on Buy Side
Figure 5b. Top 10 Stock Picks per Sector on Sell Side
- Some stocks are favored by Cramer with in the same sector: $FB ranked high on the buy side whereas $GOOGL appears on both buy and sell recommendations.
- Cramer doesn't like your “meme stocks”: $NKLA, $GME etc are all high up on his sell list.
Profit from a “buy and hold” strategy?
Plenty of folks have said to pay no attention to Cramer. But what if we applied the most basic strategy to his recommendations: Would we profit from a “buy and hold” strategy?
Among Cramer’s 927 buy side recommendations, 739 of them did not have corresponding sell mentions, that is, these were “buy-only” stock picks. If we bought these Cramer picks on the earliest dates they were recommended, what would our gains be? Also, how does this strategy compare to returns from the S&P 500?
From the 739 “buy-only” stock picks, 641 were on the NYSE so we could analyse their historical price changes and compare them to the S&P 500 (note that this analysis did not account for inflation). In Figure 6, Red bars are the “buy-only” stock picks that had negative returns, while Green bars are stock picks that had positive returns if bought in April 2016 and held to Oct 6, 2021.
While it appears 458 out of 641 (~71%) of Cramer’s “buy-only” picks returned gains, when we normalize them S&P 500 returns only 174 (~27%) beat the market (e.g. if a stock had 60% returns, and S&P 500 also returned 60% over the same period, the fold change is 1, the S&P 500 threshold is noted as a blue dotted line in Figure 6).
Figure 6. Returns on Cramer’s buy-only picks using a Buy and Hold strategy
This quick comparison suggests the simple “buy-and-hold” strategy will likely yield some returns, but beating the S&P 500 benchmark will be less likely in the long term.
Examining the “Cramer Bounce” today
Now that we’ve found that Cramer’s “buy-only” picks are unlikely to beat the market over the long term, how about the Cramer Bounce? The Cramer bounce is the observed phenomenon of a sudden overnight rise in a stock's price after it has been recommended by Cramer on Mad Money. If the Cramer bounce still exists or if there’s a temporary fluctuation caused by Cramer’s recommendations, retail traders could design strategies to make quick gains.
This report is by no means the first of its kind to investigate whether Cramer’s recommendations resulted in abnormal returns. A Google search on this topic will guide you to various outlets that have looked into this... and they have vastly different conclusions. Some have concluded that Cramer had quality stock picks and analysis but with no data. While others have analyzed Cramer’s picks and argued that they were: 1) inconclusive, 2) added value if one bought and held, or 3) in certain conditions resulted in astonishing returns. We won’t link to all the published materials on this topic, but clearly some caution is required when interpreting these studies. Some concerns we have include:
- As mentioned in the previous section, not all mentions by Cramer are equal.
- The sample sizes in some studies are very small. Taking 100 or so stocks out of his list is perhaps not the best practice.
- Sampling period should include bull and bear-ish years. One can recommend stocks blindly in a bullish year and still generate decent returns.
- The time period used to analyze Cramer’s effect on the stock is important. Some studies look at the effect one-day after his recommendation, but the performance of the previous day should also be considered. Longer term returns such as 7 days or 30 days, etc. should also be considered to match different trading styles (e.g. intraday vs buy and hold).
- Caution should be placed on the stock price used to calculate the next day's return. The show airs at 6pm each day after market close. For the majority of retail traders, the next day return should be calculated using the next day's open and close price, instead of the conventional close-to-close price.
- Not all stocks are equal. They are different in market capitalization, sector, volatility, momentum, etc. These variables should be taken into account when deciding the presence of abnormal returns. Factor analysis proposed by Bolster, Trahan and Venkateswaran is certainly a more fair approach than comparing to the simple market average.
Figure 7 shows a preliminary examination of Cramer’s stock buy recommendation and the short term daily return trends in each sector that were covered in Figures 2-5, using the time interval [1,1]. This time interval means that:
- One day before the show, left boundary “1”
- The show date (trading hours happens before the show announcement), considered time “0”
- The immediate next trading day, the right boundary “1”
All the stock prices of the 8,000+ buy recommendations on each date of announcement were first normalized for comparing different stock prices on different days from 2016 to 2021. Next, normalized prices were used to calculate the daily returns using the conventional close-to-close prices of each day, before aggregation by mean into year and sector. One caveat for using the conventional close-to-close prices is the level of return is not readily achievable for regular retail traders since the after and before market trading.
In Figure 7, we sacrificed the resolution of individual stocks and daily changes, however it should become obvious if there is a “Cramer Bounce” or any effect in different sectors and if that impact is consistent throughout the last couple years. Figure 7 shows many sectors had higher daily returns following the show date when stocks in those sectors were recommended on air. This may suggest that Cramer’s announcements likely had some correlation to that positive performance.
Figure 7. Normalized Stock Daily Returns by Sector
Interestingly, although Cramer’s buy side stock picks were in favor of technology and consumer cyclical sectors, the sectors that had the biggest “bounce” following the show were in real estate (2016, 2017), basic materials (2018), utilities (2019, 2020, 2021), basic materials (2018, 2020), and industrials (2021). His favorite recommended sectors, technology and consumer cyclical, on the contrary, remained rather undisturbed over the announcements in the 6 years analyzed.
Factoring the stock mention bias in the sectors (Figure 3), his stock recommendations perhaps did not catalyse the intended “Cramer Bounce” effect between 2016 and 2021. However, there seems to be observable change in sector-based average short term changes. This warrants a closer look.
At the end of the day, all models are wrong, but some are useful. To determine whether Cramer’s recommendations had a significant impact on the stock’s short term returns, one has to figure out how to predict the “expected” price of a stock on any given day. Only when the expected return is known, can one truly judge whether the observed return is abnormal. In reality, no one can tell what the price of a stock should be tomorrow.
Without getting too deep into the academic research and fitting over 8000+ stock mentions into fancy models, this report aimed to simply detect if there was a difference at all, not for a single stock, but for all stocks Cramer recommended after appearing on Mad Money. Although the presence of significant differences in stock returns before and after Cramer’s announcement does not prove causation, it strongly implied a prominent impact. This answers the question if Cramer, being a public influencer, can indeed influence the stock performance or not.
In Figure 8, we focused on the retail trader's actionable “daily returns” (calculated by the difference between each day’s open and close) for every buy recommendation between 2016 and 2021. The general idea is that each of the 8000+ stock picks move randomly up or down a few percentages on a given day, but if their mean “daily return” is significantly higher or lower compared to the previous dates after Cramer’s announcement, there is a high chance that Cramer’s recommendation had a real impact on the stock prices, given the one common connection for all of them is their appearance on his show.
Figure 8 examined the “daily returns” distribution between the time interval [1,1] and tested whether there were any significant differences between those days. To avoid bombarding readers with too much data, only 2016 and 2021 results are shown below.
For each figure, the small panels represent a month, and the three violin plots represent the day before the Mad Money show (Red), the day the stock was recommended (Green), and one day after (Blue). The shape of each violin gives us the distribution of “daily returns”, where the plot is the widest is where the majority of data points are located. We can see the general shapes are similar between the 3 colored violin plots in each month, this means the open-close difference between the 3 dates are similar for most of the buy recommendations, consistent with the previous results in this report and others. The differences (extreme long ends) are from a subpopulation of stocks, possibly smaller cap and growth stocks similar to what Bolster and Trahan found. The most noticeable difference could be observed in May 2016 (Figure 8, 2016, Month 5) where the show’s next day return is significantly lower than the show date and day before. At the same time, the show date daily return isn't that much different from the previous date. This is a pattern suggesting Cramer’s mention in May 2016 likely made a difference after his show’s announcement, since nothing happened before the show. Combined with where the mean lies (“wide waist”), in this case, Cramer’s mentions in May actually ended up significantly dragging the daily performance of those stocks.
Figure 8. Assessing Cramer’s mention’s impact on stock daily returns
Interestingly, for several months in the 2021, the next day returns were distributed lower than the show date’s, meaning the mean returns were actually going down following Cramer’s recommendations! In 2021, there may be a phenomenon where if Cramer mentions a stock, then it may mean bad news (at least in the short term) for that ticker. This is very different from the 2000s’ “Cramer Bounce”.
Okay, so we see from the 2 snapshots of 2016 and 2021 that Cramer’s recommendations might have some impact on the stock’s daily returns. How frequently does this happen on a monthly basis between 2016 and 2021? Is his influence consistent or decreasing? And lastly, do we see the returns mostly go up or down?
To get an overview of the situation, Figure 9 tallies the number of months where Cramer’s buy recommendations have a significant change in next day returns (Increasing return or Decreasing return). The bar chart in the bottom panel shows the total number of months in each year that experienced significant changes on daily returns. The count has been pretty steady around 6 months per year where Cramer’s recommendations had an impact on next day daily returns. The upper panel line plots show the odds of Increasing return or Decreasing return in percentage out of the total months per year. The line plot provides additional information on the trend: Cramer’s buy recommendations were more likely to coincide with a decrease in daily stock returns in more recent years (2020/2021 vs. 2016).
Figure 9. Summary Cramer’s picks and impact on daily returns 2016-2021
The same analysis can be applied to longer time intervals such as [3,3], [7,7] and [30,30]. That is, we can compare the cumulative stock returns 3/7/30 days before the show and 3/7/30 days after to detect if there’s a significant difference. Since the time period is longer, the returns are calculated by daily close prices.
An example of changes over a longer time period is shown in Figure 10. It compares stock returns in a [7, 7] time frame: 7 days before Cramer’s buy recommendations and 7 days after the announcement. We see 5 out 9 months detected a significant change. Consistent with the [1,1] analysis, most of these returns also trend downwards following the show.
Figure 10. Assessing Cramer’s picks and the impact on 2016 stock returns with a [7, 7] time frame
When we replicate Figure 9’s analysis over longer time intervals, we see that over a slightly longer period (3/7/30 days) returns typically decrease following Cramer’s recommendations (Figure 11 below). To be sure, this is a consistent feature across all time frames [1,1], [3,3], [7,7] and [30,30] where decreasing returns are more likely than increasing returns
It’s interesting to note that in 2020-21 this trend is seeing a slight reversal, which may be due to favorable market conditions rather than Cramer’s prowess in picking stocks.
Figure 11. Cramer buys and longer returns between 2016-2021
Does influence expire?
Now that we see Cramer's stock recommendations can be associated with a significant change on daily returns and likely extends for a short period of time. This is actually quite remarkable, considering the amount of stocks mentioned in each of those months. Moreover, we’ve observed a “Cramer Dive” in recent years, instead of the intended bounce in the early 2000s, which could be leveraged by retail traders.
Another side of market influence: let’s try to determine how professional fund managers and institutional analysts respond to Cramer’s stock recommendations. After all, it is not news that some of them are shorting Cramer’s buy recommendations. Is Cramer’s influence among the establishment waning?
This type of analysis has rarely been attempted, at least not in any quantitative way. Here, we attempt to measure Cramer’s influence on institutional analysts (the establishment) and gain some insight that might be useful for retail traders.
We have tried our best to gather analysts’ ratings and recommendations throughout the years on the stocks mentioned in Cramer’s show. Each available analyst’s recommendation was tracked down to the date. The analysis follows what each analyst thinks of the stocks on Cramer’s buy recommendations by examining their ratings on the same day of Cramer’s show, and up to 5 days later. Counting buy ratings out of the total analysts’ ratings on that date related to Cramer’s picks, a “Cramer Index” was calculated. The higher the index value, the stronger agreement on Cramer’s buy recommendation. For example, if 10 stocks are mentioned by Cramer on the show, 8 of them are rated by other analysts 3 days after, and 4 of the analysts also hold a positive/buy attitude, the Cramer index will be 4/8=0.5. Note the following caveats for this analysis:
- We cannot gather every rating made by every analyst, nor do they all comment on the same stocks at the same time.
- We don’t have the exact timestamp data for the analysts’ recommendations made on the Cramer show date. So analysts’ ratings made on the Cramer’s show day were not considered
Figure 12. Analysts responses to stocks in Cramer’s mentions 2016-2021
Although direct cause-and-effect relationships cannot be determined, Figure 12 still helps us to infer the establishment’s response to Cramer’s buy recommendations by looking at the value of Cramer indexes and how each colored line shifts.
Looking at the red line (1 day after the show) as a reference, between 2016 and 2017, analysts’ ratings (if made) on Cramer’s picks were highly consistent. In 2018, a shake-up occurred: 1-day and 2-day dropped to the lowest in the past 5 years, meaning analysts were giving the opposite ratings to Cramer’s buy recommendations within 2 days of the show. Although both lines rebounded quite a bit in recent years and have matched the other time intervals, all time intervals have seen a decrease in Cramer index values in recent years. This may imply a sentiment change among analysts to recommend the opposite to Cramer. This actually makes sense if we look at Figure 9’s decreasing return trend (red line), which represents the decreasing daily return 1 day after the show that has shot up since 2019.
This report cannot draw a simple conclusion on whether Jim Cramer is a stock genius. However, in both parts of the report, we can see that Cramer knows his way around stock investments: he managed to draft a handful of “market beaters” in his early days as a fund manager. What this is exactly due to, or a result of, we may never know.
As a media personality, Cramer’s stock picking can still make you some money, although if you buy and hold you are less likely to beat the market. Remarkably, we see his impact on short term stock performance still exists in recent years like in the early 2000s.
Contrary to the “Cramer bounce”, which was coined in the early 2000s, his impact on short term stock performance actually shows a “Cramer Dive” in the immediate post-show periods. This unintended consequence might have something to do with the attitude switch of fund managers and institutional analysts.
Cramer has decided to leave Mad Money and expand his role at CNBC into multi-platform media. Whether this has anything to do with his show’s waning influence on the market or with the influx of a new generation of retail traders who live on their smart devices, it is key to his future success as an influencer to stay relevant and reach more people. We wish him luck.